During this fiscal year we have made significant progress on several projects.[unreadable] [unreadable] 1. Dissecting functional domains of the leukemogenic CBF&#946;-SMMHC fusion protein with knock-in mouse lines expressing truncated versions of CBF&#946;-SMMHC.[unreadable] Core-binding factor (CBF) is a heterodimeric transcription factor composed of two subunits, RUNX1 and Cbfb. CBFB is fused with MYH11, the gene encoding smooth muscle heavy chain (SMMHC), as a result of inv(16) in cases of AML subtype M4Eo. Molecularly the fusion protein CBF&#946;-SMMHC is able to interact with RUNX1, homo-dimerize and multimerize. CBF&#946;-SMMHC can sequester RUNX1 in the cytoplasm. It has higher affinity for RUNX1 and protects RUNX1 from ubiquitin degradation better than the wild type Cbf&#946; does. In addition, certain region of SMMHC may serve as a transcription repressor. However, the in vivo functional importance of these mechanisms is not clear.[unreadable] In this study we intend to determine the relative importance of CBF&#946;-SMMHC functional domains through the knock-in approach. We generated knock in chimeric mice expressing CBF&#946;-SMMHC with 3 C-terminal and 2 internal deletions. One of the deletions removes only the domain responsible for high-affinity binding of RUNX1 (D43 mouse) and the resulting mutant was less efficient in sequestering RUNX1, and caused less severe hematopoietic defects in F1 embryos than full length CBF&#946;-SMMHC. Contrary to expectations, most chimeric mice and F1 mice expressing this deleted protein developed leukemia spontaneously shortly after birth, unlike mice with full-length CBF&#946;-SMMHC, which only develop leukemia after ENU or retroviral mutagenesis. Mice heterozygous for another larger deletion removes both the RUNX1 high-affinity binding and stabilization domains(D198 mouse) developed severe bone development defects, which are similar to but different from previous Runx2 or Cbfb models and did not show appreciable defects in hematopoiesis, and did not develop AML after ENU treatment. Mice expressing a mutant deleting the domains for multimerization, transcriptional repression (DC95 mouse) did not show any embryonic and adult hematopoietic blockage, while did develop lethal myeoproliferative disorders without ENU treatment around 1 year after birth. These data from knock-ins suggest that the RUNX1 high-affinity binding and stabilization domains, and also multimerization and repressor domain are important for CBF&#946;-SMMHC to block hematopoiesis. On the other hand, the ability of CBF&#946;-SMMHC to induce leukemia is not dependent on its ability to repress RUNX1 function. These findings have enhanced our understanding of the leukemogenic process and may provide more specific targets for novel therapeutic development.[unreadable] [unreadable] 2. Cbfb-MYH11 activities independent of Runx/Cbfb repression in primitive hematopoiesis and leukemia. [unreadable] The leukemic fusion gene CBFB-MYH11 is thought to function through dominant repression of RUNX1 and CBFB. Previous data from mouse models support this hypothesis as embryos heterozygous for Cbfb-MYH11 (Cbfb+/MYH11) have the same phenotype as Cbfb-/- and Runx1-/- embryos: failure of definitive hematopoiesis and CNS hemorrhaging. However, we also observed potential defects in primitive hematopoiesis in the Cbfb+/MYH11 embryos. As primitive blood defects have not been described in Cbfb-/- embryos, these observations raise the possibility that Cbfb-MYH11 has activities not related to Cbfb inhibition. To address this hypothesis, we further characterized primitive blood cells in Cbfb+/MYH11, Cbfb-/-, and Runx1-/- embryos. We found that Cbfb-MYH11 caused a delay in differentiation that was not seen in Cbfb-/- embryos and was more severe than in Runx1-/- embryos. We also found that primitive blood cells from Cbfb+/MYH11, but not Cbfb-/- or Runx1-/-, embryos had increased proliferation and apoptosis. Furthermore, microarray analysis demonstrated unique gene expresssion changes in the Cbfb+/MYH11 primitive blood cells. We found that three of these genes, Csfrb, Il1rl1, and Gata2 are similarly upregulated at the protein level in Cbfb-MYH11 expressing adult mice. Interestingly, all three of these genes can be linked, either directly or indirectly with eosinophil development. As eosiophilia is a common feature unique to CBFB-MYH11 leukemias, but not leukemias involving other RUNX family members, these findings imply that CBFB-MYH11 also has RUNX/CBFB independent activities which may contribute to leukemogenesis.[unreadable] [unreadable] 3. The effect of Runx1 mutations on leukemogenesis by Cbfb-SMMHC.[unreadable] As mentioned above, current hypothesis is that Runx1 binding and repression is a key step in leukemogenesis by Cbfb-SMMHC. To test this hypothesis, we have determined the rate of leukemia development in Cbfb-MYH11 knockin mice that contain one of the two mutant alleles of Runx1 Runx1+/- and Runx1+/LZ. Runx1+/- contains a null allele and Runx1+/LZ contains a fusion between RD and LacZ, deleting C-terminal transactivation domains of Runx1. This latter allele was found to be partially dominant-negative. The Runx1-repression model would predict accelerated leukemogenesis in these mice. However, our preliminary data showed that Cbfb-MYH11 mice that were Runx1+/- had similar rate of leukemogenesis compared with Cbfb-MYH11 mice that were Runx1+/+ after ENU treatment. Moreover, Cbfb-MYH11 mice that were Runx1+/LZ had delayed leukemogenesis compared to Cbfb-MYH11 mice that were Runx1+/+. These results are inconsistent with the hypothesis that Runx1 repression is the key mechanism in leukemogenesis by Cbfb-SMMHC. Alternative hypotheses include Runx1-repression independent function of Cbfb-SMMHC and RD of Runx1 acting as a repressor of Cbfb-SMMHC. Further experiments are being carried out to test these hypotheses.[unreadable] [unreadable] 4. Cooperation between CBFB-MYH11 and activated KIT gene[unreadable] KIT mutations are common in AML with inv(16), with reported frequencies at 10-45%. The most common mutations are amino acid substitutions at codon D816, which result in constitutive activation of the encoded KIT tyrosine kinase. Clinically patients with both inv(16) and KIT mutations have worse prognosis than those with inv(16) alone. Therefore, it is important to verify the cooperation between KIT mutation and Cbfb-MYH11, and investigate the underlying mechanism in leukemogenesis. This may provide a novel insight for future drug development. [unreadable] We transduced retroviral vectors expressing KIT with D816V or D816Y mutations into bone marrow cells harvested from mice with the Cbfb-MYH11 gene. The transduced bone marrow cells were transplanted to sub-lethally irradiated mice, which were then monitored for leukemia development. Within five months after transplantation, 60% of mice carrying D816Y and 50% of mice carrying D816V KIT mutations have died from AML, while all GFP controls were alive. This data suggested a strong cooperation between KIT D816 mutations and Cbfb-MYH11 during leukemogenesis. In vitro colony formation assay experiments showed that Cbfb-MYH11 bone marrow cells could form more colonies than wild type controls. With the addition of KIT D816Y mutations, the percentage of myeloid coloies was increased. This data suggested that the KIT mutations drove cells towards myeloid lineage. This mouse model will be used to determine the combined therapeutic effects of inhibitors against both CBF&#946;-SMMHC and activated KIT.